{
"slug": "consensus-watch-monthly-2026-09",
"series": "monthly",
"series_label": "MONTHLY · CONSENSUS WATCH",
"issue": 2,
"title": "Monthly Consensus Watch #2",
"language": "en",
"window_start": "2026-09-01T00:00:00+00:00",
"window_end": "2026-10-01T00:00:00+00:00",
"window_text": "Sep 1-30, 2026 UTC",
"cutoff": "2026-10-01T16:00:00+00:00",
"generated_at": "2026-10-02T09:18:46+00:00",
"methodology": "v1.1 (2026-08-10)",
"methodology_hash": "e66c7e8c864a2233",
"source": "monthly_metrics_2026-09.json",
"research_question": "if a model's forecast matches the consensus of the other models, is the forecast more reliable?",
"method": "leave-one-out strict majority, >=3 directional peers",
"base_field_hit": {
"n": 20682,
"hit_rate": 0.455
},
"market_state": {
"panel": "market_state_6p2",
"month": {
"start": "2026-09-01",
"end": "2026-10-01",
"days": 30
},
"prev_month_partial": {
"start": "2026-08-01",
"end": "2026-09-01",
"days": 31,
"partial": false
},
"current": {
"btc_net_pct": 6.42,
"realized_vol_ann_pct": 41.17,
"volume_top5_usd_bn": 80.0142,
"volume_top5_mom_pct": 21.3,
"avg_pairwise_corr": 0.78
},
"previous_partial_month": {
"btc_net_pct": 24.95,
"realized_vol_ann_pct": 41.53,
"volume_top5_usd_bn": 65.9696,
"avg_pairwise_corr": 0.79
},
"snapshot_line": "BTC net +6.42% (prior +24.95%) - ann vol 41.2% (was 41.5%) - TOP5 volume $80.01B, +21% m/m - pairwise corr 0.78 (was 0.79)",
"exchange": "binance",
"audit_filter_applied": true,
"source": "single-exchange (binance) daily candle recompute, market_state_2026-09.json; the prior-month block is the pack's own prev_month (Aug 1-31)",
"audit_note": "Market-state row is computed on a single-exchange (binance) daily candle series, as in the weekly series after the audit that found the raw table mixes two exchanges. The prior-month block is the full August: it covers 31 days (Aug 1-31), and the TOP5 volume figure compares a 30-day sum with a 31-day sum -- a level difference, not a like-for-like change.",
"weeks": {
"W1": {
"btc_net_pct": 3.42,
"realized_vol_ann_pct": 43.35,
"regime": "calm"
},
"W2": {
"btc_net_pct": -4.36,
"realized_vol_ann_pct": 19.68,
"regime": "calm"
},
"W3": {
"btc_net_pct": 5.64,
"realized_vol_ann_pct": 51.01,
"regime": "volatile"
},
"W4": {
"btc_net_pct": 4.06,
"realized_vol_ann_pct": 52.25,
"regime": "volatile"
}
},
"volatility_rule": "volatile if realized_vol_ann_pct >= median(W1..W4) else calm",
"median_realized_vol_ann_pct": 47.18,
"delta_month_vs_prev_partial": {
"btc_net_pct": -18.53,
"realized_vol_ann_pct": -0.36,
"avg_daily_range_pct": 0.06,
"volume_top5_usd_pct": 21.3,
"avg_pairwise_corr": -0.01,
"note": "September (30 days) vs August (31 days, the full previous month); volume is a sum, so volume_top5_usd_pct compares a 30-day sum with a 31-day sum (not normalised per day)"
},
"controls": {
"w4_matches_published": {
"pins_source": "market_state_2026-09-21.json -> 'current' (the published weekly #8 week, Sep 21-27)",
"block": "W4",
"published": {
"btc_net_pct": 4.06,
"realized_vol_ann_pct": 52.25,
"volume_top5_usd_bn": 22.4265,
"avg_pairwise_corr": 0.71
},
"got": {
"btc_net_pct": 4.06,
"realized_vol_ann_pct": 52.25,
"volume_top5_usd_bn": 22.4265,
"avg_pairwise_corr": 0.71
},
"match": {
"btc_net_pct": true,
"realized_vol_ann_pct": true,
"volume_top5_usd_bn": true,
"avg_pairwise_corr": true
},
"tolerance": {
"btc_net_pct": 0.05,
"realized_vol_ann_pct": 0.5,
"volume_top5_usd_bn": 0.05,
"avg_pairwise_corr": 0.02
},
"all_match": true
},
"w3_matches_published_previous": {
"pins_source": "market_state_2026-09-21.json -> 'previous' (the weekly #8 run's prev week = weekly #7, Sep 14-20)",
"block": "W3",
"published": {
"btc_net_pct": 5.64,
"realized_vol_ann_pct": 51.01,
"volume_top5_usd_bn": 18.7386,
"avg_pairwise_corr": 0.93
},
"got": {
"btc_net_pct": 5.64,
"realized_vol_ann_pct": 51.01,
"volume_top5_usd_bn": 18.7386,
"avg_pairwise_corr": 0.93
},
"match": {
"btc_net_pct": true,
"realized_vol_ann_pct": true,
"volume_top5_usd_bn": true,
"avg_pairwise_corr": true
},
"tolerance": {
"btc_net_pct": 0.05,
"realized_vol_ann_pct": 0.5,
"volume_top5_usd_bn": 0.05,
"avg_pairwise_corr": 0.02
},
"all_match": true
},
"note": "W4 and W3 of this run are exactly the two weeks the weekly #8 market panel computed, with the same audit filter, so both must reproduce. Non-fatal: a MISMATCH means the exchange filter or the symbol set moved -- send the log, do not publish the row."
},
"regime_days": {
"2026-09-01": {
"move_pct": -1.5,
"regime": "trend"
},
"2026-09-02": {
"move_pct": -0.06,
"regime": "flat"
},
"2026-09-03": {
"move_pct": 4.98,
"regime": "trend"
},
"2026-09-04": {
"move_pct": -1.97,
"regime": "trend"
},
"2026-09-05": {
"move_pct": 0.24,
"regime": "flat"
},
"2026-09-06": {
"move_pct": 0.58,
"regime": "flat"
},
"2026-09-07": {
"move_pct": -1.52,
"regime": "trend"
},
"2026-09-08": {
"move_pct": -0.72,
"regime": "flat"
},
"2026-09-09": {
"move_pct": -0.25,
"regime": "flat"
},
"2026-09-10": {
"move_pct": -2.14,
"regime": "trend"
},
"2026-09-11": {
"move_pct": 0.85,
"regime": "flat"
},
"2026-09-12": {
"move_pct": 0.0,
"regime": "flat"
},
"2026-09-13": {
"move_pct": -0.57,
"regime": "flat"
},
"2026-09-14": {
"move_pct": 1.79,
"regime": "trend"
},
"2026-09-15": {
"move_pct": -3.35,
"regime": "trend"
},
"2026-09-16": {
"move_pct": 0.86,
"regime": "flat"
},
"2026-09-17": {
"move_pct": 0.21,
"regime": "flat"
},
"2026-09-18": {
"move_pct": 5.82,
"regime": "trend"
},
"2026-09-19": {
"move_pct": 0.46,
"regime": "flat"
},
"2026-09-20": {
"move_pct": 0.13,
"regime": "flat"
},
"2026-09-21": {
"move_pct": 6.39,
"regime": "trend"
},
"2026-09-22": {
"move_pct": -0.4,
"regime": "flat"
},
"2026-09-23": {
"move_pct": -2.04,
"regime": "trend"
},
"2026-09-24": {
"move_pct": 0.02,
"regime": "flat"
},
"2026-09-25": {
"move_pct": -0.46,
"regime": "flat"
},
"2026-09-26": {
"move_pct": 0.36,
"regime": "flat"
},
"2026-09-27": {
"move_pct": 0.2,
"regime": "flat"
},
"2026-09-28": {
"move_pct": -1.19,
"regime": "flat"
},
"2026-09-29": {
"move_pct": 0.22,
"regime": "flat"
},
"2026-09-30": {
"move_pct": -0.22,
"regime": "flat"
}
},
"alive_slice_context": {
"abs_move_1d_pct": {
"A": 1.979,
"B": 1.905
},
"btc_realized_vol_ann_pct_hourly": {
"A": 36.92,
"B": 34.88
}
}
},
"tables": {
"pooled": {
"agree": {
"n": 17488,
"hit_rate": 0.4507,
"wilson_95": [
0.4433,
0.4581
]
},
"disagree": {
"n": 137,
"hit_rate": 0.5474,
"wilson_95": [
0.4639,
0.6284
]
},
"lift_pp": -9.7,
"thin_or_tied": 3057,
"by_margin": {
"2": {
"n": 87,
"hit_rate": 0.4253
},
"3": {
"n": 1712,
"hit_rate": 0.4585
},
"4": {
"n": 2400,
"hit_rate": 0.4971
},
"5": {
"n": 3972,
"hit_rate": 0.4383
},
"6": {
"n": 9317,
"hit_rate": 0.4428
}
},
"total_scoreable": 17625,
"caveat": "The disagree bucket is n=137 over the whole month; the two Wilson intervals do not overlap. Prior window Aug 1-31: agree 45.5% / disagree 43.6%, lift +1.9pp, on n=15,878 agree and n=211 disagree. These are dependent observations pooled across the whole month: two calm weeks, two volatile weeks and 3 days outside them."
},
"by_cell": [
{
"agree": {
"n": 10928,
"hit_rate": 0.4693,
"wilson_95": [
0.4599,
0.4786
]
},
"disagree": {
"n": 106,
"hit_rate": 0.5472,
"wilson_95": [
0.4524,
0.6386
]
},
"lift_pp": -7.8,
"thin_or_tied": 1974,
"by_margin": {
"2": {
"n": 54,
"hit_rate": 0.4074
},
"3": {
"n": 1184,
"hit_rate": 0.4688
},
"4": {
"n": 1660,
"hit_rate": 0.4958
},
"5": {
"n": 2472,
"hit_rate": 0.4454
},
"6": {
"n": 5558,
"hit_rate": 0.4727
}
},
"n_subjects": 13008,
"cell": "fh_1h_tf_1h",
"fh": "1h",
"tf": "1h"
},
{
"agree": {
"n": 2683,
"hit_rate": 0.4439,
"wilson_95": [
0.4252,
0.4628
]
},
"disagree": {
"n": 11,
"hit_rate": 0.6364,
"wilson_95": [
0.3538,
0.8483
]
},
"lift_pp": -19.2,
"thin_or_tied": 445,
"by_margin": {
"2": {
"n": 21,
"hit_rate": 0.4286
},
"3": {
"n": 208,
"hit_rate": 0.4712
},
"4": {
"n": 270,
"hit_rate": 0.563
},
"5": {
"n": 588,
"hit_rate": 0.4371
},
"6": {
"n": 1596,
"hit_rate": 0.4229
}
},
"n_subjects": 3139,
"cell": "fh_4h_tf_4h",
"fh": "4h",
"tf": "4h"
},
{
"agree": {
"n": 2826,
"hit_rate": 0.4243,
"wilson_95": [
0.4062,
0.4426
]
},
"disagree": {
"n": 15,
"hit_rate": 0.5333,
"wilson_95": [
0.3012,
0.7519
]
},
"lift_pp": -10.9,
"thin_or_tied": 481,
"by_margin": {
"2": {
"n": 9,
"hit_rate": 0.3333
},
"3": {
"n": 252,
"hit_rate": 0.4286
},
"4": {
"n": 350,
"hit_rate": 0.5229
},
"5": {
"n": 696,
"hit_rate": 0.4325
},
"6": {
"n": 1519,
"hit_rate": 0.3976
}
},
"n_subjects": 3322,
"cell": "fh_4h_tf_1h",
"fh": "4h",
"tf": "1h"
},
{
"agree": {
"n": 493,
"hit_rate": 0.3529,
"wilson_95": [
0.312,
0.3961
]
},
"disagree": {
"n": 0,
"hit_rate": null,
"wilson_95": [
null,
null
]
},
"lift_pp": null,
"thin_or_tied": 55,
"by_margin": {
"3": {
"n": 40,
"hit_rate": 0.4
},
"4": {
"n": 60,
"hit_rate": 0.25
},
"5": {
"n": 120,
"hit_rate": 0.35
},
"6": {
"n": 273,
"hit_rate": 0.37
}
},
"n_subjects": 548,
"cell": "fh_1d_tf_4h",
"fh": "1d",
"tf": "4h"
},
{
"agree": {
"n": 558,
"hit_rate": 0.3405,
"wilson_95": [
0.3024,
0.3808
]
},
"disagree": {
"n": 5,
"hit_rate": 0.4,
"wilson_95": [
0.1176,
0.7693
]
},
"lift_pp": -5.9,
"thin_or_tied": 102,
"by_margin": {
"2": {
"n": 3,
"hit_rate": 1.0
},
"3": {
"n": 28,
"hit_rate": 0.2857
},
"4": {
"n": 60,
"hit_rate": 0.3333
},
"5": {
"n": 96,
"hit_rate": 0.4167
},
"6": {
"n": 371,
"hit_rate": 0.3208
}
},
"n_subjects": 665,
"cell": "fh_1d_tf_1d",
"fh": "1d",
"tf": "1d"
}
],
"by_margin": [
{
"margin": 2,
"n": 87,
"hit_rate": 0.4253
},
{
"margin": 3,
"n": 1712,
"hit_rate": 0.4585
},
{
"margin": 4,
"n": 2400,
"hit_rate": 0.4971
},
{
"margin": 5,
"n": 3972,
"hit_rate": 0.4383
},
{
"margin": 6,
"n": 9317,
"hit_rate": 0.4428
}
],
"by_margin_month_over_month": [
{
"margin": 3,
"prior_window_hit_rate": 0.4593,
"prior_window_n": 1964,
"monthly2_hit_rate": 0.4585,
"monthly2_n": 1712,
"prior_source": "month.prior_window.consensus.by_margin",
"prior_note": null
},
{
"margin": 4,
"prior_window_hit_rate": 0.4513,
"prior_window_n": 2630,
"monthly2_hit_rate": 0.4971,
"monthly2_n": 2400,
"prior_source": "month.prior_window.consensus.by_margin",
"prior_note": null
},
{
"margin": 5,
"prior_window_hit_rate": 0.4581,
"prior_window_n": 4482,
"monthly2_hit_rate": 0.4383,
"monthly2_n": 3972,
"prior_source": "month.prior_window.consensus.by_margin",
"prior_note": null
},
{
"margin": 6,
"prior_window_hit_rate": 0.4511,
"prior_window_n": 6664,
"monthly2_hit_rate": 0.4428,
"monthly2_n": 9317,
"prior_source": "consensus-watch-monthly-2026-08.json tables.pooled.by_margin[6] (the published Monthly #1, window Aug 1-31)",
"prior_note": null
}
],
"per_model": [
{
"agree": {
"n": 2587,
"hit_rate": 0.455,
"wilson_95": [
0.4359,
0.4742
]
},
"disagree": {
"n": 6,
"hit_rate": 0.1667,
"wilson_95": [
0.0301,
0.5635
]
},
"lift_pp": 28.8,
"thin_or_tied": 352,
"by_margin": {
"2": {
"n": 10,
"hit_rate": 0.3
},
"3": {
"n": 240,
"hit_rate": 0.4667
},
"4": {
"n": 367,
"hit_rate": 0.5068
},
"5": {
"n": 639,
"hit_rate": 0.446
},
"6": {
"n": 1331,
"hit_rate": 0.444
}
},
"model": "claude-fable-5"
},
{
"agree": {
"n": 2347,
"hit_rate": 0.4533,
"wilson_95": [
0.4333,
0.4735
]
},
"disagree": {
"n": 4,
"hit_rate": 0.25,
"wilson_95": [
0.0456,
0.6994
]
},
"lift_pp": 20.3,
"thin_or_tied": 181,
"by_margin": {
"2": {
"n": 11,
"hit_rate": 0.3636
},
"3": {
"n": 174,
"hit_rate": 0.4713
},
"4": {
"n": 284,
"hit_rate": 0.4789
},
"5": {
"n": 547,
"hit_rate": 0.4333
},
"6": {
"n": 1331,
"hit_rate": 0.4545
}
},
"model": "claude-opus-5"
},
{
"agree": {
"n": 2623,
"hit_rate": 0.4468,
"wilson_95": [
0.4279,
0.4659
]
},
"disagree": {
"n": 30,
"hit_rate": 0.5667,
"wilson_95": [
0.392,
0.7262
]
},
"lift_pp": -12.0,
"thin_or_tied": 699,
"by_margin": {
"2": {
"n": 16,
"hit_rate": 0.5
},
"3": {
"n": 292,
"hit_rate": 0.4658
},
"4": {
"n": 396,
"hit_rate": 0.5
},
"5": {
"n": 588,
"hit_rate": 0.449
},
"6": {
"n": 1331,
"hit_rate": 0.4252
}
},
"model": "deepseek-v4-pro"
},
{
"agree": {
"n": 2522,
"hit_rate": 0.4722,
"wilson_95": [
0.4528,
0.4918
]
},
"disagree": {
"n": 72,
"hit_rate": 0.625,
"wilson_95": [
0.5095,
0.7278
]
},
"lift_pp": -15.3,
"thin_or_tied": 690,
"by_margin": {
"2": {
"n": 14,
"hit_rate": 0.4286
},
"3": {
"n": 298,
"hit_rate": 0.4933
},
"4": {
"n": 354,
"hit_rate": 0.5339
},
"5": {
"n": 525,
"hit_rate": 0.4552
},
"6": {
"n": 1331,
"hit_rate": 0.4583
}
},
"model": "gemini-3.1-pro"
},
{
"agree": {
"n": 2660,
"hit_rate": 0.4515,
"wilson_95": [
0.4327,
0.4705
]
},
"disagree": {
"n": 7,
"hit_rate": 0.5714,
"wilson_95": [
0.2505,
0.8418
]
},
"lift_pp": -12.0,
"thin_or_tied": 466,
"by_margin": {
"2": {
"n": 11,
"hit_rate": 0.4545
},
"3": {
"n": 285,
"hit_rate": 0.4281
},
"4": {
"n": 394,
"hit_rate": 0.4822
},
"5": {
"n": 639,
"hit_rate": 0.4429
},
"6": {
"n": 1331,
"hit_rate": 0.4515
}
},
"model": "gpt-5.6-sol"
},
{
"agree": {
"n": 2097,
"hit_rate": 0.4354,
"wilson_95": [
0.4143,
0.4567
]
},
"disagree": {
"n": 14,
"hit_rate": 0.3571,
"wilson_95": [
0.1634,
0.6124
]
},
"lift_pp": 7.8,
"thin_or_tied": 160,
"by_margin": {
"2": {
"n": 11,
"hit_rate": 0.4545
},
"3": {
"n": 122,
"hit_rate": 0.4098
},
"4": {
"n": 220,
"hit_rate": 0.4773
},
"5": {
"n": 413,
"hit_rate": 0.414
},
"6": {
"n": 1331,
"hit_rate": 0.4373
}
},
"model": "grok-4.6"
},
{
"agree": {
"n": 2652,
"hit_rate": 0.4389,
"wilson_95": [
0.4201,
0.4579
]
},
"disagree": {
"n": 4,
"hit_rate": 0.5,
"wilson_95": [
0.15,
0.85
]
},
"lift_pp": -6.1,
"thin_or_tied": 509,
"by_margin": {
"2": {
"n": 14,
"hit_rate": 0.4286
},
"3": {
"n": 301,
"hit_rate": 0.4518
},
"4": {
"n": 385,
"hit_rate": 0.4909
},
"5": {
"n": 621,
"hit_rate": 0.4219
},
"6": {
"n": 1331,
"hit_rate": 0.429
}
},
"model": "qwen-3.8-max"
}
],
"herd_worst_cards": [
{
"slot": "2026-09-22T00:01:00+00:00",
"symbol": "SOL",
"fh": "1d",
"tf": "1d",
"side": "long",
"n_models": 7,
"mean_conf": 70.4,
"hit_rate": 0.0
},
{
"slot": "2026-09-26T00:01:00+00:00",
"symbol": "SOL",
"fh": "1d",
"tf": "1d",
"side": "long",
"n_models": 7,
"mean_conf": 70.4,
"hit_rate": 0.0
},
{
"slot": "2026-09-18T20:01:00+00:00",
"symbol": "XRP",
"fh": "4h",
"tf": "4h",
"side": "long",
"n_models": 7,
"mean_conf": 70,
"hit_rate": 0.0
},
{
"slot": "2026-09-21T01:01:00+00:00",
"symbol": "SOL",
"fh": "1h",
"tf": "1h",
"side": "long",
"n_models": 7,
"mean_conf": 69.7,
"hit_rate": 0.1429
},
{
"slot": "2026-09-16T09:01:00+00:00",
"symbol": "BNB",
"fh": "1h",
"tf": "1h",
"side": "short",
"n_models": 7,
"mean_conf": 69.6,
"hit_rate": 0.0
}
],
"counters": {
"forecasts_total": 44100,
"ok_in_gate": 39806,
"out_of_gate": 4185,
"out_of_gate_by_fh": {
"1w": 2096,
"1m": 2089
},
"invalid": 109,
"mature": 39806,
"mature_directional": 20682,
"mature_sideways": 19124,
"pending_next_issue": 0,
"pending_by_fh": {},
"late_closes": 0,
"uptime": [
{
"fh": "1h",
"tf": "1h",
"slots_seen": 720,
"slots_expected": 720
},
{
"fh": "4h",
"tf": "4h",
"slots_seen": 180,
"slots_expected": 180
},
{
"fh": "4h",
"tf": "1h",
"slots_seen": 180,
"slots_expected": 180
},
{
"fh": "1d",
"tf": "1d",
"slots_seen": 30,
"slots_expected": 30
},
{
"fh": "1d",
"tf": "4h",
"slots_seen": 30,
"slots_expected": 30
}
],
"regime_days": {
"2026-09-01": {
"move_pct": -1.5,
"regime": "trend"
},
"2026-09-02": {
"move_pct": -0.06,
"regime": "flat"
},
"2026-09-03": {
"move_pct": 4.98,
"regime": "trend"
},
"2026-09-04": {
"move_pct": -1.97,
"regime": "trend"
},
"2026-09-05": {
"move_pct": 0.24,
"regime": "flat"
},
"2026-09-06": {
"move_pct": 0.58,
"regime": "flat"
},
"2026-09-07": {
"move_pct": -1.52,
"regime": "trend"
},
"2026-09-08": {
"move_pct": -0.72,
"regime": "flat"
},
"2026-09-09": {
"move_pct": -0.25,
"regime": "flat"
},
"2026-09-10": {
"move_pct": -2.14,
"regime": "trend"
},
"2026-09-11": {
"move_pct": 0.85,
"regime": "flat"
},
"2026-09-12": {
"move_pct": 0.0,
"regime": "flat"
},
"2026-09-13": {
"move_pct": -0.57,
"regime": "flat"
},
"2026-09-14": {
"move_pct": 1.79,
"regime": "trend"
},
"2026-09-15": {
"move_pct": -3.35,
"regime": "trend"
},
"2026-09-16": {
"move_pct": 0.86,
"regime": "flat"
},
"2026-09-17": {
"move_pct": 0.21,
"regime": "flat"
},
"2026-09-18": {
"move_pct": 5.82,
"regime": "trend"
},
"2026-09-19": {
"move_pct": 0.46,
"regime": "flat"
},
"2026-09-20": {
"move_pct": 0.13,
"regime": "flat"
},
"2026-09-21": {
"move_pct": 6.39,
"regime": "trend"
},
"2026-09-22": {
"move_pct": -0.4,
"regime": "flat"
},
"2026-09-23": {
"move_pct": -2.04,
"regime": "trend"
},
"2026-09-24": {
"move_pct": 0.02,
"regime": "flat"
},
"2026-09-25": {
"move_pct": -0.46,
"regime": "flat"
},
"2026-09-26": {
"move_pct": 0.36,
"regime": "flat"
},
"2026-09-27": {
"move_pct": 0.2,
"regime": "flat"
},
"2026-09-28": {
"move_pct": -1.19,
"regime": "flat"
},
"2026-09-29": {
"move_pct": 0.22,
"regime": "flat"
},
"2026-09-30": {
"move_pct": -0.22,
"regime": "flat"
}
},
"grok_era": {
"ids_in_window": [
"grok-4.6"
],
"n_grok45": 0,
"n_grok46": 6300,
"n_grok45_rows": 0,
"n_grok46_rows": 6300,
"first_grok46_slot": "2026-09-01T00:01:00+00:00",
"last_grok45_slot": null,
"first_grok45_slot": null,
"flip_at": "2026-08-24T09:22:00+00:00",
"flip_inside_window": false,
"prev_window": {
"n_grok45_rows": 4520,
"n_grok46_rows": 1570
},
"merged_model_id": "grok-4.6",
"merged_label": "grok-4.6",
"lineage_ids": [
"grok-4.5",
"grok-4.6"
],
"lineage_source": "benchmarks.config.lineage_ids(head) for each of the 7 heads + MODEL_SUCCESSION[claude-opus-5=('claude-opus-4.8',), grok-4.6=('grok-4.5',), qwen-3.8-max=('qwen-3.7-max',)]",
"lineage_warnings": [],
"rows_remapped_a_plus_b": 4520,
"note": "window MONTH [2026-09-01T00:00:00+00:00, 2026-10-01T00:00:00+00:00): grok-4.5 rows 0, grok-4.6 rows 6300; flip 2026-08-24T09:22:00+00:00 is NOT inside this window. All per-model aggregates use the lineage splice grok-4.5 -> grok-4.6 (one row, labelled 'grok-4.6'). See lineage_era for the same facts across all 7 heads.",
"cells_with_two_grok_rows": 0
},
"lineage_era": {
"window_key": "MONTH",
"heads": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"lineage_map": {
"claude-opus-4.8": "claude-opus-5",
"grok-4.5": "grok-4.6",
"qwen-3.7-max": "qwen-3.8-max"
},
"legacy_by_head": {
"claude-fable-5": [],
"claude-opus-5": [
"claude-opus-4.8"
],
"deepseek-v4-pro": [],
"gemini-3.1-pro": [],
"gpt-5.6-sol": [],
"grok-4.6": [
"grok-4.5"
],
"qwen-3.8-max": [
"qwen-3.7-max"
]
},
"lineage_source": "benchmarks.config.lineage_ids(head) for each of the 7 heads + MODEL_SUCCESSION[claude-opus-5=('claude-opus-4.8',), grok-4.6=('grok-4.5',), qwen-3.8-max=('qwen-3.7-max',)]",
"lineage_resolved_heads": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"lineage_warnings": [],
"per_head": {
"claude-fable-5": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "claude-fable-5"
},
"claude-opus-5": {
"legacy_ids": [
"claude-opus-4.8"
],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"n_rows_remapped_prev_window": {
"claude-opus-4.8": 0
},
"by_legacy_id": {
"claude-opus-4.8": {
"n_rows_remapped": {
"MONTH": 0,
"prev_window": 0
},
"first_slot": null,
"last_slot": null,
"first_slot_prev_window": null,
"last_slot_prev_window": null
}
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "claude-opus-5"
},
"deepseek-v4-pro": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "deepseek-v4-pro"
},
"gemini-3.1-pro": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "gemini-3.1-pro"
},
"gpt-5.6-sol": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "gpt-5.6-sol"
},
"grok-4.6": {
"legacy_ids": [
"grok-4.5"
],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"n_rows_remapped_prev_window": {
"grok-4.5": 4520
},
"by_legacy_id": {
"grok-4.5": {
"n_rows_remapped": {
"MONTH": 0,
"prev_window": 4520
},
"first_slot": null,
"last_slot": null,
"first_slot_prev_window": "2026-08-01T00:01:00+00:00",
"last_slot_prev_window": "2026-08-24T09:01:00+00:00"
}
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "grok-4.6"
},
"qwen-3.8-max": {
"legacy_ids": [
"qwen-3.7-max"
],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"n_rows_remapped_prev_window": {
"qwen-3.7-max": 870
},
"by_legacy_id": {
"qwen-3.7-max": {
"n_rows_remapped": {
"MONTH": 0,
"prev_window": 870
},
"first_slot": null,
"last_slot": null,
"first_slot_prev_window": "2026-08-01T00:01:00+00:00",
"last_slot_prev_window": "2026-08-05T09:01:00+00:00"
}
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "qwen-3.8-max"
}
},
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"rows_remapped_window": 0,
"rows_remapped_prev_window": 5390,
"note": "Only the 7 current models are tracked; earlier versions are folded into their successors' lineage. Every per-model aggregate in this pack sees the merged head id; the raw id survives in model_raw. Labels and era spans are generated from the pulled rows, not hardcoded.",
"cells_with_two_rows_per_head": {
"claude-fable-5": 0,
"claude-opus-5": 0,
"deepseek-v4-pro": 0,
"gemini-3.1-pro": 0,
"gpt-5.6-sol": 0,
"grok-4.6": 0,
"qwen-3.8-max": 0
}
},
"series_density": {
"1d": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": true,
"n_slots": 30,
"n_rows": 2100,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"1h": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": true,
"n_slots": 720,
"n_rows": 25200,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"1m": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": false,
"n_slots": 30,
"n_rows": 2100,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"1w": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": false,
"n_slots": 30,
"n_rows": 2100,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"4h": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": true,
"n_slots": 180,
"n_rows": 12600,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
}
},
"model_roster": {
"current": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"current_count": 7,
"current_raw_ids_in_window": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"archived_legacy": [
"grok-4.5",
"qwen-3.7-max",
"claude-opus-4.8"
],
"archived_legacy_count": 3,
"archived_legacy_detail": [
{
"model": "grok-4.5",
"n_rows_since": 11128,
"first_slot": "2026-07-15T18:01:00+00:00",
"last_slot": "2026-08-24T09:01:00+00:00"
},
{
"model": "qwen-3.7-max",
"n_rows_since": 7479,
"first_slot": "2026-07-15T18:01:00+00:00",
"last_slot": "2026-08-05T09:01:00+00:00"
},
{
"model": "claude-opus-4.8",
"n_rows_since": 6315,
"first_slot": "2026-07-15T18:01:00+00:00",
"last_slot": "2026-07-30T08:01:00+00:00"
}
],
"archived_excluding_lineage_merged": [],
"tracked_total": 10,
"canon_expected": {
"current": 7,
"archived_legacy": 4,
"tracked_total": 11
},
"matches_canon_11_7_4": false,
"roster_rule_w213": "7 models tracked (current line-up; earlier versions folded into their successors' lineage)",
"canon_expected_w213": {
"current": 7
},
"heads_w213": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"matches_canon_7_current": true,
"heads_missing_in_window": [],
"heads_unexpected_in_window": [],
"roster_since": "2026-07-11T00:00:00+00:00",
"note": "current = merged-lineage model ids emitting inside the window (legacy ids are spliced into their successor); archived legacy = raw ids seen since 2026-07-11 that no longer emit in the window. Models retired before 2026-07-11 are not visible to this query. The canon checks are DESCRIPTIVE: if the data says otherwise, the data wins and is printed."
}
},
"by_week": {
"rows": [
{
"week": "W1",
"label": "Sep 1-6",
"published_label": "Aug 31-Sep 6",
"n_scoreable": 3191,
"herd_share_pct": 99.3,
"agree_hit": 0.4328,
"disagree_hit": 0.6957,
"lift_pp": -26.3,
"unanimity_cells": 333,
"unanimity_n": 2214,
"unanimity_hit": 0.4033,
"published_lift_pp": -27.6,
"published_n_scoreable": 3751,
"published_unanimous_cells_ge6": 392
},
{
"week": "W2",
"label": "Sep 7-13",
"published_label": "Sep 7-13",
"n_scoreable": 3788,
"herd_share_pct": 99.3,
"agree_hit": 0.3878,
"disagree_hit": 0.6923,
"lift_pp": -30.4,
"unanimity_cells": 417,
"unanimity_n": 2761,
"unanimity_hit": 0.3665,
"published_lift_pp": -30.4,
"published_n_scoreable": 3788,
"published_unanimous_cells_ge6": 417
},
{
"week": "W3",
"label": "Sep 14-20",
"published_label": "Sep 14-20",
"n_scoreable": 4877,
"herd_share_pct": 98.9,
"agree_hit": 0.5059,
"disagree_hit": 0.3846,
"lift_pp": 12.1,
"unanimity_cells": 539,
"unanimity_n": 3617,
"unanimity_hit": 0.4963,
"published_lift_pp": 12.1,
"published_n_scoreable": 4877,
"published_unanimous_cells_ge6": 539
},
{
"week": "W4",
"label": "Sep 21-27",
"published_label": "Sep 21-27",
"n_scoreable": 4245,
"herd_share_pct": 99.3,
"agree_hit": 0.4598,
"disagree_hit": 0.6,
"lift_pp": -14.0,
"unanimity_cells": 505,
"unanimity_n": 3393,
"unanimity_hit": 0.4766,
"published_lift_pp": -14.0,
"published_n_scoreable": 4245,
"published_unanimous_cells_ge6": 505
},
{
"week": "month",
"label": "Sep 1-30",
"n_scoreable": 17625,
"herd_share_pct": 99.2,
"agree_hit": 0.4507,
"disagree_hit": 0.5474,
"lift_pp": -9.7,
"unanimity_cells": 1965,
"unanimity_n": 13121,
"unanimity_hit": 0.4423
}
],
"source": "month_pull_split_by_slot_week (cutoff 2026-10-01T16:00:00+00:00)",
"unanimity_definition": "unanimity = a cell (slot x symbol x fh x tf) in which every directional forecast is on the same side and at least 6 models are present -- the same N the w201 herd block uses (counters/consensus.herd.unanimous_cells_ge6); the LOO peer floor is 3 peers.",
"published_cross_check": {
"W1": {
"n_scoreable": 3751,
"agree_n": 3725,
"agree_hit": 0.4158,
"agree_wilson_95": [
0.4001,
0.4317
],
"disagree_n": 26,
"disagree_hit": 0.6923,
"disagree_wilson_95": [
0.5001,
0.835
],
"lift_pp": -27.6,
"herd_share_pct": 99.3,
"thin_or_tied": 769,
"by_margin_all": {
"2": {
"n": 21,
"hit_rate": 0.2381
},
"3": {
"n": 456,
"hit_rate": 0.4474
},
"4": {
"n": 615,
"hit_rate": 0.5398
},
"5": {
"n": 918,
"hit_rate": 0.4303
},
"6": {
"n": 1715,
"hit_rate": 0.3574
}
},
"unanimous_cells_ge6": 392,
"source": "weeks.W1.consensus.loo_pooled (frozen cutoff)"
},
"W2": {
"n_scoreable": 3788,
"agree_n": 3762,
"agree_hit": 0.3878,
"agree_wilson_95": [
0.3724,
0.4035
],
"disagree_n": 26,
"disagree_hit": 0.6923,
"disagree_wilson_95": [
0.5001,
0.835
],
"lift_pp": -30.4,
"herd_share_pct": 99.3,
"thin_or_tied": 680,
"by_margin_all": {
"2": {
"n": 15,
"hit_rate": 0.8
},
"3": {
"n": 400,
"hit_rate": 0.4275
},
"4": {
"n": 550,
"hit_rate": 0.4691
},
"5": {
"n": 984,
"hit_rate": 0.4035
},
"6": {
"n": 1813,
"hit_rate": 0.3425
}
},
"unanimous_cells_ge6": 417,
"source": "weeks.W2.consensus.loo_pooled (frozen cutoff)"
},
"W3": {
"n_scoreable": 4877,
"agree_n": 4825,
"agree_hit": 0.5059,
"agree_wilson_95": [
0.4918,
0.52
],
"disagree_n": 52,
"disagree_hit": 0.3846,
"disagree_wilson_95": [
0.2647,
0.5204
],
"lift_pp": 12.1,
"herd_share_pct": 98.9,
"thin_or_tied": 721,
"by_margin_all": {
"2": {
"n": 45,
"hit_rate": 0.4222
},
"3": {
"n": 424,
"hit_rate": 0.5212
},
"4": {
"n": 685,
"hit_rate": 0.5547
},
"5": {
"n": 990,
"hit_rate": 0.4424
},
"6": {
"n": 2681,
"hit_rate": 0.5159
}
},
"unanimous_cells_ge6": 539,
"source": "weeks.W3.consensus.loo_pooled (frozen cutoff)"
},
"W4": {
"n_scoreable": 4245,
"agree_n": 4215,
"agree_hit": 0.4598,
"agree_wilson_95": [
0.4448,
0.4749
],
"disagree_n": 30,
"disagree_hit": 0.6,
"disagree_wilson_95": [
0.4232,
0.7541
],
"lift_pp": -14.0,
"herd_share_pct": 99.3,
"thin_or_tied": 713,
"by_margin_all": {
"2": {
"n": 9,
"hit_rate": 0.6667
},
"3": {
"n": 304,
"hit_rate": 0.3914
},
"4": {
"n": 485,
"hit_rate": 0.3897
},
"5": {
"n": 876,
"hit_rate": 0.468
},
"6": {
"n": 2541,
"hit_rate": 0.4778
}
},
"unanimous_cells_ge6": 505,
"source": "weeks.W4.consensus.loo_pooled (frozen cutoff)"
}
},
"weeks_matching_published_lift": [
"W2",
"W3",
"W4"
]
},
"margin_by_week": {
"tiers": [
"3",
"4",
"5",
"6"
],
"rows": [
{
"margin": 3,
"weeks": {
"W1": {
"n": 384,
"hit_rate": 0.4844
},
"W2": {
"n": 400,
"hit_rate": 0.4275
},
"W3": {
"n": 424,
"hit_rate": 0.5212
},
"W4": {
"n": 304,
"hit_rate": 0.3914
}
},
"month": {
"n": 1712,
"hit_rate": 0.4585
}
},
{
"margin": 4,
"weeks": {
"W1": {
"n": 525,
"hit_rate": 0.5333
},
"W2": {
"n": 550,
"hit_rate": 0.4691
},
"W3": {
"n": 685,
"hit_rate": 0.5547
},
"W4": {
"n": 485,
"hit_rate": 0.3897
}
},
"month": {
"n": 2400,
"hit_rate": 0.4971
}
},
{
"margin": 5,
"weeks": {
"W1": {
"n": 732,
"hit_rate": 0.4822
},
"W2": {
"n": 984,
"hit_rate": 0.4035
},
"W3": {
"n": 990,
"hit_rate": 0.4424
},
"W4": {
"n": 876,
"hit_rate": 0.468
}
},
"month": {
"n": 3972,
"hit_rate": 0.4383
}
},
{
"margin": 6,
"weeks": {
"W1": {
"n": 1512,
"hit_rate": 0.3651
},
"W2": {
"n": 1813,
"hit_rate": 0.3425
},
"W3": {
"n": 2681,
"hit_rate": 0.5159
},
"W4": {
"n": 2541,
"hit_rate": 0.4778
}
},
"month": {
"n": 9317,
"hit_rate": 0.4428
}
}
],
"source": "month_pull_split_by_slot_week (cutoff 2026-10-01T16:00:00+00:00)"
},
"prior_window_consensus": {
"loo_pooled": {
"n_scoreable": 16089,
"agree_n": 15878,
"agree_hit": 0.4554,
"agree_wilson_95": [
0.4477,
0.4632
],
"disagree_n": 211,
"disagree_hit": 0.436,
"disagree_wilson_95": [
0.3708,
0.5035
],
"lift_pp": 1.9,
"herd_share_pct": 98.7,
"excluded_thin_tied": 3294,
"thin_or_tied": 3294,
"min_peers": 3
},
"by_margin": [
{
"peers": 3,
"n": 1964,
"hit_rate": 0.4593,
"wilson_95": [
0.4373,
0.4814
]
},
{
"peers": 4,
"n": 2630,
"hit_rate": 0.4513,
"wilson_95": [
0.4324,
0.4704
]
},
{
"peers": 5,
"n": 4482,
"hit_rate": 0.4581,
"wilson_95": [
0.4435,
0.4727
]
},
{
"peers": "unanimity",
"n": 10984,
"hit_rate": 0.4539,
"wilson_95": [
0.4446,
0.4633
],
"cells": 1672,
"min_models": 6
}
],
"loo_by_cell": {
"fh_1h_tf_1h": {
"n_scoreable": 9982,
"agree_n": 9857,
"agree_hit": 0.4519,
"disagree_n": 125,
"disagree_hit": 0.48,
"lift_pp": -2.8,
"n_subjects": 12110
},
"fh_4h_tf_1h": {
"n_scoreable": 2573,
"agree_n": 2548,
"agree_hit": 0.449,
"disagree_n": 25,
"disagree_hit": 0.4,
"lift_pp": 4.9,
"n_subjects": 3039
},
"fh_4h_tf_4h": {
"n_scoreable": 2537,
"agree_n": 2483,
"agree_hit": 0.4611,
"disagree_n": 54,
"disagree_hit": 0.3889,
"lift_pp": 7.2,
"n_subjects": 3086
},
"fh_1d_tf_1d": {
"n_scoreable": 459,
"agree_n": 454,
"agree_hit": 0.5396,
"disagree_n": 5,
"disagree_hit": 0.2,
"lift_pp": 34.0,
"n_subjects": 539
},
"fh_1d_tf_4h": {
"n_scoreable": 538,
"agree_n": 536,
"agree_hit": 0.4534,
"disagree_n": 2,
"disagree_hit": 0.0,
"lift_pp": 45.3,
"n_subjects": 609
}
},
"loo_by_model": [
{
"n_scoreable": 2404,
"agree_n": 2391,
"agree_hit": 0.458,
"disagree_n": 13,
"disagree_hit": 0.5385,
"lift_pp": -8.0,
"model": "claude-fable-5",
"insufficient": false
},
{
"n_scoreable": 2109,
"agree_n": 2105,
"agree_hit": 0.4727,
"disagree_n": 4,
"disagree_hit": 0.5,
"lift_pp": -2.7,
"model": "claude-opus-5",
"insufficient": false
},
{
"n_scoreable": 2104,
"agree_n": 2065,
"agree_hit": 0.4489,
"disagree_n": 39,
"disagree_hit": 0.4103,
"lift_pp": 3.9,
"model": "deepseek-v4-pro",
"insufficient": false
},
{
"n_scoreable": 2338,
"agree_n": 2249,
"agree_hit": 0.4544,
"disagree_n": 89,
"disagree_hit": 0.4719,
"lift_pp": -1.7,
"model": "gemini-3.1-pro",
"insufficient": false
},
{
"n_scoreable": 2467,
"agree_n": 2459,
"agree_hit": 0.4453,
"disagree_n": 8,
"disagree_hit": 0.375,
"lift_pp": 7.0,
"model": "gpt-5.6-sol",
"insufficient": false
},
{
"n_scoreable": 2261,
"agree_n": 2218,
"agree_hit": 0.4599,
"disagree_n": 43,
"disagree_hit": 0.4419,
"lift_pp": 1.8,
"model": "grok-4.6",
"insufficient": false
},
{
"n_scoreable": 2406,
"agree_n": 2391,
"agree_hit": 0.4504,
"disagree_n": 15,
"disagree_hit": 0.2,
"lift_pp": 25.0,
"model": "qwen-3.8-max",
"insufficient": false
}
],
"unanimity": {
"cells": 1672,
"n_calls": 10984,
"hit_rate": 0.4539,
"wilson_95": [
0.4446,
0.4633
],
"insufficient": false,
"min_models": 6
},
"unanimity_definition": "unanimity = a cell (slot x symbol x fh x tf) in which every directional forecast is on the same side and at least 6 models are present -- the same N the w201 herd block uses (counters/consensus.herd.unanimous_cells_ge6); the LOO peer floor is 3 peers.",
"n_cells": 5865,
"definitions": {
"n_scoreable": "agree_n + disagree_n (subjects with >= min_peers directional peers, no tie)",
"excluded_thin_tied": "subjects dropped because the cell had < min_peers peers or a tied majority",
"lift_pp": "agree_hit - disagree_hit, percentage points",
"herd_share_pct": "agree_n / n_scoreable * 100",
"by_margin": "peers = peers on the majority side (capped at 6 as in w201); 'unanimity' = the w201 herd rule"
}
},
"verdict_block": {
"rule": "Verdict rule: each row states one reading of the month; a week agrees when it shows the same reading (same sign, or the same name). Confirmed = 3 or 4 of the 4 weeks agree with the month; Inconclusive = 2; Rejected = 0 or 1. Weekly cells come from the frozen weekly packs (7-line roster); the month is the pooled Sep 1-30 pull. W1 straddles the month edge (Aug 31).",
"header": [
"Metric",
"W1 Aug 31-Sep 6",
"W2 Sep 7-13",
"W3 Sep 14-20",
"W4 Sep 21-27",
"Month",
"Verdict"
],
"widths_mm": [
36,
26.5,
22,
22,
22,
18,
27.5
],
"weeks_source": "frozen weekly packs (weeks.W1..W4 of the monthly pack, 7-line roster)",
"month_source": "the pooled Sep 1-30 pull (month block)",
"rows": [
{
"metric": "LOO lift, pp",
"weeks": {
"W1": {
"value": "-27.6",
"reading": "negative"
},
"W2": {
"value": "-30.4",
"reading": "negative"
},
"W3": {
"value": "+12.1",
"reading": "positive"
},
"W4": {
"value": "-14.0",
"reading": "negative"
}
},
"month": {
"value": "-9.7",
"reading": "negative"
},
"agree_count": 3,
"verdict": "Confirmed (3/4)"
},
{
"metric": "Herd share >= 95%",
"weeks": {
"W1": {
"value": "99.3%",
"reading": ">= 95%"
},
"W2": {
"value": "99.3%",
"reading": ">= 95%"
},
"W3": {
"value": "98.9%",
"reading": ">= 95%"
},
"W4": {
"value": "99.3%",
"reading": ">= 95%"
}
},
"month": {
"value": "99.2%",
"reading": ">= 95%"
},
"agree_count": 4,
"verdict": "Confirmed (4/4)"
},
{
"metric": "Unanimity hit vs field hit",
"weeks": {
"W1": {
"value": "-7.1",
"reading": "negative"
},
"W2": {
"value": "-5.3",
"reading": "negative"
},
"W3": {
"value": "+0.8",
"reading": "positive"
},
"W4": {
"value": "+1.0",
"reading": "positive"
}
},
"month": {
"value": "-1.2",
"reading": "negative"
},
"agree_count": 2,
"verdict": "Inconclusive (2/4)"
}
]
}
},
"herd_note": "1,965 cells had 6+ directional models on one side; unanimous_cells_ge6 per monthly metrics",
"smart_note": "Smart-vs-Outlook deferred: as-of weights pipeline not yet recording; will appear once weights are logged per-slot (methodology 3.2.3 fallback)",
"key_finding": {
"label": "KEY FINDING · OBSERVATION (one monthly window)",
"text": "Herding ran at 99.2% of leave-one-out scoreable calls across the month and the pooled lift came in at -9.7pp: agree-hit 45.1% against 54.7% on disagreement, on a disagree bucket of n=137.",
"evidence_level": "observation"
},
"key_findings": [
"**OBSERVATION -- 99.2% of the field's leave-one-out calls ran with the herd over the month.** Of 17,625 LOO-scoreable directional calls, 17,488 (99.2%) sided with peers' majority; agree-hit was 45.1% [44.3%, 45.8%] vs. 54.7% [46.4%, 62.8%] for the 137 disagreers, a pooled lift of -9.7pp. Descriptive, one month, not a claim of a durable edge.",
"**The margin curve read 45.9% / 49.7% / 43.8% / 44.3% from 3 peers to unanimity.** 2 of the 4 tiers sit above the 45.5% field base. The same four tiers read 45.9% / 45.1% / 45.8% / 45.1% in the prior window Aug 1-31, on n=1,964 / 2,630 / 4,482 / 6,664 there.",
"**Week by week the lift changed sign 2 times.** The four published weeks ran Aug 31-Sep 6 -27.6pp, Sep 7-13 -30.4pp, Sep 14-20 +12.1pp, Sep 21-27 -14.0pp; the month, which also counts the 3 days outside those weeks, pools at -9.7pp and the prior window read +1.9pp. The month is one pooled pull, not the mean of the weeks: the weekly range is -30.4pp to +12.1pp.",
"**1,965 cells had 6+ directional models on one side across the month.** The 5 worst sat at 69.6-70.4 mean confidence and resolved 0.0%-14.3%; the prior window Aug 1-31 had 1,672 such cells, hitting 45.4% on n=10,984."
],
"market_check": {
"label": "MARKET CHECK · OBSERVATION (September against August)",
"text": "OBSERVATION — mean |1d move| 1.98% -> 1.91% (-4% rel), BTC realized vol 36.9% -> 34.9% (ann., hourly); field directional accuracy 46.3% -> 45.5% (-0.8pp), 1 of 7 models improved, sim win-rate up for 6 of 7, field sim PnL -$1054.79 -> -$2554.15 (Aug 1-31 vs Sep 1-30; descriptive, one pair of windows, not a claim).",
"robustness": "Robustness: raw price-sign accuracy 46.0% -> 48.4% — the opposite direction to the trade-based hit rule.",
"weeks": {
"A": {
"days": "2026-08-01..2026-08-31",
"slots": [
"2026-08-01T00:00:00+00:00",
"2026-09-01T00:00:00+00:00"
],
"cutoff": "2026-09-01T16:00:00+00:00",
"partial": false
},
"B": {
"days": "2026-09-01..2026-09-30",
"slots": [
"2026-09-01T00:00:00+00:00",
"2026-10-01T00:00:00+00:00"
],
"cutoff": "2026-10-01T16:00:00+00:00"
}
},
"price_source": "crypto_spot/1h@:00",
"accuracy_field": {
"A": {
"n": 19383,
"hit_rate": 0.4633,
"wilson_95": [
0.4563,
0.4703
]
},
"B": {
"n": 20682,
"hit_rate": 0.455,
"wilson_95": [
0.4483,
0.4618
]
},
"delta_pp": -0.8
},
"accuracy_per_model": [
{
"model": "gemini-3.1-pro",
"model_label": "gemini-3.1-pro",
"A": {
"n": 3033,
"hit_rate": 0.4761,
"wilson_95": [
0.4584,
0.4939
]
},
"B": {
"n": 3284,
"hit_rate": 0.4851,
"wilson_95": [
0.468,
0.5022
]
},
"delta_pp": 0.9
},
{
"model": "deepseek-v4-pro",
"model_label": "deepseek-v4-pro",
"A": {
"n": 2578,
"hit_rate": 0.4554,
"wilson_95": [
0.4362,
0.4747
]
},
"B": {
"n": 3352,
"hit_rate": 0.4526,
"wilson_95": [
0.4358,
0.4695
]
},
"delta_pp": -0.3
},
{
"model": "grok-4.6",
"model_label": "grok-4.6",
"A": {
"n": 2708,
"hit_rate": 0.4527,
"wilson_95": [
0.4341,
0.4715
]
},
"B": {
"n": 2271,
"hit_rate": 0.4439,
"wilson_95": [
0.4235,
0.4644
]
},
"delta_pp": -0.9
},
{
"model": "qwen-3.8-max",
"model_label": "qwen-3.8-max",
"A": {
"n": 3001,
"hit_rate": 0.4565,
"wilson_95": [
0.4388,
0.4744
]
},
"B": {
"n": 3165,
"hit_rate": 0.4471,
"wilson_95": [
0.4298,
0.4645
]
},
"delta_pp": -0.9
},
{
"model": "claude-fable-5",
"model_label": "claude-fable-5",
"A": {
"n": 2812,
"hit_rate": 0.4673,
"wilson_95": [
0.4489,
0.4858
]
},
"B": {
"n": 2945,
"hit_rate": 0.4537,
"wilson_95": [
0.4357,
0.4717
]
},
"delta_pp": -1.4
},
{
"model": "gpt-5.6-sol",
"model_label": "gpt-5.6-sol",
"A": {
"n": 2970,
"hit_rate": 0.4586,
"wilson_95": [
0.4407,
0.4765
]
},
"B": {
"n": 3133,
"hit_rate": 0.4446,
"wilson_95": [
0.4273,
0.4621
]
},
"delta_pp": -1.4
},
{
"model": "claude-opus-5",
"model_label": "claude-opus-5",
"A": {
"n": 2281,
"hit_rate": 0.4779,
"wilson_95": [
0.4574,
0.4984
]
},
"B": {
"n": 2532,
"hit_rate": 0.4538,
"wilson_95": [
0.4345,
0.4732
]
},
"delta_pp": -2.4
}
],
"raw_sign_field": {
"A": {
"n_scored": 19304,
"hit_rate": 0.4599
},
"B": {
"n_scored": 20654,
"hit_rate": 0.4837
}
},
"abs_move_1d": {
"A": 1.979,
"B": 1.905,
"delta_pct_rel": -3.7
},
"btc_realized_vol_ann_pct_hourly": {
"A": 36.92,
"B": 34.88
},
"trading_field": {
"A": {
"n_trades": 19383,
"wr": 0.3576,
"pnl_net_usd": -1054.79,
"pnl_gross_usd": 883.51
},
"B": {
"n_trades": 20682,
"wr": 0.384,
"pnl_net_usd": -2554.15,
"pnl_gross_usd": -485.95
}
},
"verdict": {
"models_with_hit_improved": "1 of 7",
"models_with_rawsign_improved": "7 of 7",
"models_with_wr_improved": "6 of 7",
"field_hit_delta_pp": -0.8,
"field_pnl_net": {
"A": -1054.79,
"B": -2554.15
},
"abs_move_delta": {
"1h": {
"A": 0.325,
"B": 0.356,
"delta_pct_rel": 9.5
},
"4h": {
"A": 0.724,
"B": 0.737,
"delta_pct_rel": 1.8
},
"1d": {
"A": 1.979,
"B": 1.905,
"delta_pct_rel": -3.7
}
},
"btc_vol_ann_pct": {
"A": 36.92,
"B": 34.88
},
"note": "descriptive, two adjacent windows; hit rule = methodology v1.1 trade-based; raw_sign = price-sign robustness check (closes at :00 vs slots at :01)"
},
"lineage_note": "lineage splice ACTIVE for all 7 heads: [('claude-opus-4.8', 'claude-opus-5'), ('grok-4.5', 'grok-4.6'), ('qwen-3.7-max', 'qwen-3.8-max')]. Window MONTH remapped 0 row(s), the previous window 5390. Labels are generated from the pulled rows (see lineage_era).",
"prior_window_partial": false
},
"why_it_matters": "MarketMania publishes forecasts from 7 frontier model lines into the same slots every hour; a natural question for anyone reading the feed is whether a model's call is worth more when it agrees with what everyone else is saying, or whether that agreement is just everyone making the same mistake together. The four weekly issues set against this month answered it with lifts of Aug 31-Sep 6 -27.6pp, Sep 7-13 -30.4pp, Sep 14-20 +12.1pp, Sep 21-27 -14.0pp, and the pooled month reads -9.7pp. This monthly issue asks the same question at monthly n, across four weeks the pack labels as 2 calm and 2 volatile, using the same leave-one-out design built to avoid comparing a model to an index that includes its own vote.",
"how_to_read_this": "For every model M and every forecast it made, we rebuild that slot's consensus using only the OTHER active models in the same symbol / forecast-horizon (FH) / timeframe (TF) cell -- M's own call never counts toward its own consensus. M is scored agree if its side (long or short) matches the strict majority of those peers, and disagree if it sits alone against that majority; a cell needs at least 3 directional peers to count at all, and tied peer splits are excluded rather than forced either way. LOO peers are genuinely external to the model being scored, so the comparison below is not comparing a model to itself. No model line is lineage-spliced this month (the prior window carries the grok and qwen splices), and a spliced line never counts as two peers in a cell. The month is one pull at one cutoff (Thu Oct 1, 2026, 16:00 UTC); the four weekly blocks quoted alongside it are the published issues at their own frozen cutoffs.",
"practical_implications": [
"2 of the 4 margin tiers sit above the 45.5% field base this month; the four tier rates span 43.8% to 49.7%, a spread of +5.9pp, and the four weeks set against the month ran lifts of -30.4pp to +12.1pp against a pooled monthly lift of -9.7pp.",
"The pooled monthly lift is -9.7pp while the four weeks ran -30.4pp to +12.1pp and the prior window read +1.9pp.",
"The field base moved -0.8pp from August to September (46.3% -> 45.5%), and every consensus tier is measured against that base."
],
"limitations": [
"The prior window is the full August: every \"prior\" column, the gray series of the page-1 chart and the market-check A column describe Aug 1-31. It is window A of the same pull, slots 2026-08-01 to 2026-09-01 (first slot actually seen 2026-08-01 00:01 UTC), matured at the prior cutoff Tue Sep 1, 2026, 16:00 UTC under the same engine, the same FH gate and the same hit rule as September -- 19,383 directional calls against 20,682 in the month. Month-over-month deltas are a step between two windows of different length (31 and 30 days).",
"Two cutoffs live in this issue. The month is frozen at Thu Oct 1, 2026, 16:00 UTC; the four weekly blocks are frozen at their own published cutoffs (W1 Aug 31-Sep 6, W2 Sep 7-13, W3 Sep 14-20, W4 Sep 21-27). A forecast still pending at its week's cutoff is immature in the weekly pack and mature in the month pull, so the month does not equal the sum of the weeks (20,682 vs 19,544 directional, plus 1,826 on the 3 remainder days); W1 also opens on Aug 31, a day the month pull does not hold, so its pack counts 4,520 directional calls against 3,832 for Sep 1-6 inside the month pull -- that gap is arithmetic, not a data problem.",
"Observations are not independent: the same models watch overlapping symbol / FH / TF cells hour after hour for 30 days. Wilson intervals here are descriptive, not inferential, and pooling a month does not make them inferential.",
"The pooled lift rests on 137 disagreeing calls; per-model counts run 4-72. 3,057 calls (14.8% of the 20,682-call mature directional universe) were thin or tied and are excluded from every consensus table. Correlational, not causal: LOO removes self-match bias, not the confound that an easy slot lifts agreement and hit together.",
"This window ran 10 trend days of 30, and the pack's own volatility rule splits the four Mon-Sun weeks from Aug 31 to Sep 27 as W1 volatile, W2 calm, W3 volatile, W4 calm; every month-over-month comparison is a comparison across regimes as well as across windows. Series density and lineage notes (monthly #2): (1) rolling-1w forecasts emit daily since Aug 22, 2026 and rolling-1M daily since Aug 23, 2026, so across this 30-day window daily coverage is FULL as measured -- the 1w series has slots on 30 of 30 days and the 1M series on 30 of 30 days; both sit outside this report's FH gate (1h/4h/1d) and appear only in the exclusions counter (1w 2,096, 1M 2,089). (2) No model line is lineage-spliced inside this window. (3) Every per-model aggregate in this issue sees the merged head id; the raw id is kept in the pack, and 0 rows were remapped inside the month against 5,390 in the prior window.",
"Market-state row is computed on a single-exchange (binance) daily candle series, as in the weekly series after the audit that found the raw table mixes two exchanges. The prior-month block is the full August: it covers 31 days (Aug 1-31), and the TOP5 volume figure compares a 30-day sum with a 31-day sum -- a level difference, not a like-for-like change.",
"Research-to-date counter: directional forecasts resolved inside the FH gate (1h/4h/1d) since Jul 11, counted at the issue cutoff (Thu Oct 1, 2026, 16:00 UTC) -- the D2 definition pinned in the weekly series; this issue prints 59,755. The pack carries no prior-cutoff pin for a monthly window (the prior-cutoff pin is null by construction), so the control this issue is the weekly reproduction: the W4 block of this pack (Sep 21-27) reproduces the published weekly #8 on 7 of 7 checks, and the prior window of this pack (Aug 1-31, prior cutoff Tue Sep 1, 2026, 16:00 UTC) reproduces the published Monthly #1 on 7 of 7 checks. Counter deltas against pre-wave-3 issues remain definitional."
],
"notes": {
"series_density_and_lineage_monthly2": "Series density and lineage notes (monthly #2): (1) rolling-1w forecasts emit daily since Aug 22, 2026 and rolling-1M daily since Aug 23, 2026, so across this 30-day window daily coverage is FULL as measured -- the 1w series has slots on 30 of 30 days and the 1M series on 30 of 30 days; both sit outside this report's FH gate (1h/4h/1d) and appear only in the exclusions counter (1w 2,096, 1M 2,089). (2) No model line is lineage-spliced inside this window. (3) Every per-model aggregate in this issue sees the merged head id; the raw id is kept in the pack, and 0 rows were remapped inside the month against 5,390 in the prior window.",
"grok_era": {
"ids_in_window": [
"grok-4.6"
],
"n_grok45": 0,
"n_grok46": 6300,
"n_grok45_rows": 0,
"n_grok46_rows": 6300,
"first_grok46_slot": "2026-09-01T00:01:00+00:00",
"last_grok45_slot": null,
"first_grok45_slot": null,
"flip_at": "2026-08-24T09:22:00+00:00",
"flip_inside_window": false,
"prev_window": {
"n_grok45_rows": 4520,
"n_grok46_rows": 1570
},
"merged_model_id": "grok-4.6",
"merged_label": "grok-4.6",
"lineage_ids": [
"grok-4.5",
"grok-4.6"
],
"lineage_source": "benchmarks.config.lineage_ids(head) for each of the 7 heads + MODEL_SUCCESSION[claude-opus-5=('claude-opus-4.8',), grok-4.6=('grok-4.5',), qwen-3.8-max=('qwen-3.7-max',)]",
"lineage_warnings": [],
"rows_remapped_a_plus_b": 4520,
"note": "window MONTH [2026-09-01T00:00:00+00:00, 2026-10-01T00:00:00+00:00): grok-4.5 rows 0, grok-4.6 rows 6300; flip 2026-08-24T09:22:00+00:00 is NOT inside this window. All per-model aggregates use the lineage splice grok-4.5 -> grok-4.6 (one row, labelled 'grok-4.6'). See lineage_era for the same facts across all 7 heads.",
"cells_with_two_grok_rows": 0
},
"model_ids_line": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"model_roster": {
"current": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"current_count": 7,
"current_raw_ids_in_window": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"archived_legacy": [
"grok-4.5",
"qwen-3.7-max",
"claude-opus-4.8"
],
"archived_legacy_count": 3,
"archived_legacy_detail": [
{
"model": "grok-4.5",
"n_rows_since": 11128,
"first_slot": "2026-07-15T18:01:00+00:00",
"last_slot": "2026-08-24T09:01:00+00:00"
},
{
"model": "qwen-3.7-max",
"n_rows_since": 7479,
"first_slot": "2026-07-15T18:01:00+00:00",
"last_slot": "2026-08-05T09:01:00+00:00"
},
{
"model": "claude-opus-4.8",
"n_rows_since": 6315,
"first_slot": "2026-07-15T18:01:00+00:00",
"last_slot": "2026-07-30T08:01:00+00:00"
}
],
"archived_excluding_lineage_merged": [],
"tracked_total": 10,
"canon_expected": {
"current": 7,
"archived_legacy": 4,
"tracked_total": 11
},
"matches_canon_11_7_4": false,
"roster_rule_w213": "7 models tracked (current line-up; earlier versions folded into their successors' lineage)",
"canon_expected_w213": {
"current": 7
},
"heads_w213": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"matches_canon_7_current": true,
"heads_missing_in_window": [],
"heads_unexpected_in_window": [],
"roster_since": "2026-07-11T00:00:00+00:00",
"note": "current = merged-lineage model ids emitting inside the window (legacy ids are spliced into their successor); archived legacy = raw ids seen since 2026-07-11 that no longer emit in the window. Models retired before 2026-07-11 are not visible to this query. The canon checks are DESCRIPTIVE: if the data says otherwise, the data wins and is printed."
},
"series_density": {
"1d": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": true,
"n_slots": 30,
"n_rows": 2100,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"1h": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": true,
"n_slots": 720,
"n_rows": 25200,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"1m": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": false,
"n_slots": 30,
"n_rows": 2100,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"1w": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": false,
"n_slots": 30,
"n_rows": 2100,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"4h": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": true,
"n_slots": 180,
"n_rows": 12600,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
}
},
"month_over_month_note": "Month-over-month columns compare back-to-back windows: Aug 1-31 vs Sep 1-30 (this issue). \"Was\" values are the pack's own prior-window block -- window A of the same pull, same engine, same FH gate and same hit rule, matured at the prior cutoff Tue Sep 1, 2026, 16:00 UTC. The prior month is the full August (31 days against 30 in September), and it holds 19,383 directional calls against 20,682 in September. Deltas are computed on unrounded rates.",
"research_to_date_pin": "Research-to-date counter: directional forecasts resolved inside the FH gate (1h/4h/1d) since Jul 11, counted at the issue cutoff (Thu Oct 1, 2026, 16:00 UTC) -- the D2 definition pinned in the weekly series; this issue prints 59,755. The pack carries no prior-cutoff pin for a monthly window (the prior-cutoff pin is null by construction), so the control this issue is the weekly reproduction: the W4 block of this pack (Sep 21-27) reproduces the published weekly #8 on 7 of 7 checks, and the prior window of this pack (Aug 1-31, prior cutoff Tue Sep 1, 2026, 16:00 UTC) reproduces the published Monthly #1 on 7 of 7 checks. Counter deltas against pre-wave-3 issues remain definitional.",
"market_state_audit": "Market-state row is computed on a single-exchange (binance) daily candle series, as in the weekly series after the audit that found the raw table mixes two exchanges. The prior-month block is the full August: it covers 31 days (Aug 1-31), and the TOP5 volume figure compares a 30-day sum with a 31-day sum -- a level difference, not a like-for-like change.",
"engine": "Engine 1.1 has powered the sandbox since Aug 18, i.e. before this window and inside the prior window; no cross-engine PnL comparisons are claimed.",
"report_template": "PDF built with the canonical mm_report.py template module; preview card with report_preview.render_preview",
"lineage_era": {
"window_key": "MONTH",
"heads": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"lineage_map": {
"claude-opus-4.8": "claude-opus-5",
"grok-4.5": "grok-4.6",
"qwen-3.7-max": "qwen-3.8-max"
},
"legacy_by_head": {
"claude-fable-5": [],
"claude-opus-5": [
"claude-opus-4.8"
],
"deepseek-v4-pro": [],
"gemini-3.1-pro": [],
"gpt-5.6-sol": [],
"grok-4.6": [
"grok-4.5"
],
"qwen-3.8-max": [
"qwen-3.7-max"
]
},
"lineage_source": "benchmarks.config.lineage_ids(head) for each of the 7 heads + MODEL_SUCCESSION[claude-opus-5=('claude-opus-4.8',), grok-4.6=('grok-4.5',), qwen-3.8-max=('qwen-3.7-max',)]",
"lineage_resolved_heads": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"lineage_warnings": [],
"per_head": {
"claude-fable-5": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "claude-fable-5"
},
"claude-opus-5": {
"legacy_ids": [
"claude-opus-4.8"
],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"n_rows_remapped_prev_window": {
"claude-opus-4.8": 0
},
"by_legacy_id": {
"claude-opus-4.8": {
"n_rows_remapped": {
"MONTH": 0,
"prev_window": 0
},
"first_slot": null,
"last_slot": null,
"first_slot_prev_window": null,
"last_slot_prev_window": null
}
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "claude-opus-5"
},
"deepseek-v4-pro": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "deepseek-v4-pro"
},
"gemini-3.1-pro": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "gemini-3.1-pro"
},
"gpt-5.6-sol": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "gpt-5.6-sol"
},
"grok-4.6": {
"legacy_ids": [
"grok-4.5"
],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"n_rows_remapped_prev_window": {
"grok-4.5": 4520
},
"by_legacy_id": {
"grok-4.5": {
"n_rows_remapped": {
"MONTH": 0,
"prev_window": 4520
},
"first_slot": null,
"last_slot": null,
"first_slot_prev_window": "2026-08-01T00:01:00+00:00",
"last_slot_prev_window": "2026-08-24T09:01:00+00:00"
}
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "grok-4.6"
},
"qwen-3.8-max": {
"legacy_ids": [
"qwen-3.7-max"
],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"n_rows_remapped_prev_window": {
"qwen-3.7-max": 870
},
"by_legacy_id": {
"qwen-3.7-max": {
"n_rows_remapped": {
"MONTH": 0,
"prev_window": 870
},
"first_slot": null,
"last_slot": null,
"first_slot_prev_window": "2026-08-01T00:01:00+00:00",
"last_slot_prev_window": "2026-08-05T09:01:00+00:00"
}
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "qwen-3.8-max"
}
},
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"rows_remapped_window": 0,
"rows_remapped_prev_window": 5390,
"note": "Only the 7 current models are tracked; earlier versions are folded into their successors' lineage. Every per-model aggregate in this pack sees the merged head id; the raw id survives in model_raw. Labels and era spans are generated from the pulled rows, not hardcoded.",
"cells_with_two_rows_per_head": {
"claude-fable-5": 0,
"claude-opus-5": 0,
"deepseek-v4-pro": 0,
"gemini-3.1-pro": 0,
"gpt-5.6-sol": 0,
"grok-4.6": 0,
"qwen-3.8-max": 0
}
},
"frozen_cutoffs": "Two cutoffs live in this issue. The month is frozen at Thu Oct 1, 2026, 16:00 UTC; the four weekly blocks are frozen at their own published cutoffs (W1 Aug 31-Sep 6, W2 Sep 7-13, W3 Sep 14-20, W4 Sep 21-27). A forecast still pending at its week's cutoff is immature in the weekly pack and mature in the month pull, so the month does not equal the sum of the weeks (20,682 vs 19,544 directional, plus 1,826 on the 3 remainder days); W1 also opens on Aug 31, a day the month pull does not hold, so its pack counts 4,520 directional calls against 3,832 for Sep 1-6 inside the month pull -- that gap is arithmetic, not a data problem.",
"partial_prior_window": "The prior window is the full August: every \"prior\" column, the gray series of the page-1 chart and the market-check A column describe Aug 1-31. It is window A of the same pull, slots 2026-08-01 to 2026-09-01 (first slot actually seen 2026-08-01 00:01 UTC), matured at the prior cutoff Tue Sep 1, 2026, 16:00 UTC under the same engine, the same FH gate and the same hit rule as September -- 19,383 directional calls against 20,682 in the month. Month-over-month deltas are a step between two windows of different length (31 and 30 days)."
},
"bibtex_key": "mm_consensus_monthly_2026m09",
"citation": {
"bibtex_key": "mm_consensus_monthly_2026m09",
"title": "Monthly Consensus Watch #2: does agreeing with the crowd make an LLM's market call safer?",
"author": "MarketMania Research",
"year": 2026,
"month": "October",
"day": 2,
"url": "https://marketmania.ai/research/reports/consensus-watch-monthly-2026-09.pdf",
"note": "Methodology v1.1; window Sep 1-30, 2026 UTC; source monthly_metrics_2026-09.json"
},
"living_series_note": "**Issue #2.** Monthly Consensus Watch is a living monthly comparison; each issue appends one more calendar month of leave-one-out agreement-vs-hit data. This second issue reads September against the full August: herd share 99.2%, pooled lift -9.7pp (agree 45.1% vs disagree 54.7% on n=137), four weekly lifts set against the month running -30.4pp to +12.1pp, and the prior window Aug 1-31 at herd share 98.7% and lift +1.9pp. No model line is lineage-spliced this month (the prior window carries the grok and qwen splices). Engine 1.1 has powered the sandbox since Aug 18, i.e. before this window and inside the prior window; no cross-engine PnL comparisons are claimed.",
"testing_next": [
"Monthly #3 (October): the same three series over Oct 1-31, cut on Nov 1 plus the maturity lag, so its month-over-month comparison runs against this September.",
"Monthly Config Watch #2: the same September window through the config search engine (w298), published as the fourth monthly report of this issue.",
"Quarterly Stability Index: Run 3, the next repeat-prompt stability run, is planned for the first half of October and is the first cross-quarter point on model stability."
],
"related_research": [
{
"title": "Monthly Calibration #2",
"url": "https://marketmania.ai/research/reports/calibration-monthly-2026-09.pdf"
},
{
"title": "Monthly Model Watch #2",
"url": "https://marketmania.ai/research/reports/model-watch-monthly-2026-09.pdf"
},
{
"title": "Consensus Watch #8",
"url": "https://marketmania.ai/research/reports/consensus-watch-2026-09-21.pdf"
}
],
"related_research_note": "The three monthly reports publish together as one issue each month; each links straight to the others' PDF and to the last weekly issue of the same series. Direct links are the posting rule from wave 2 on.",
"research_to_date": {
"this_report": {
"loo_scored_observations": 17625
},
"platform": {
"as_of_cutoff": "2026-10-01",
"resolved_forecasts": 59755,
"resolved_forecasts_definition": "directional forecasts resolved inside the FH gate (1h/4h/1d) since Jul 11, at the issue cutoff",
"resolved_forecasts_since": "2026-07-11",
"models_tracked": 7,
"models_tracked_detail": "7 models tracked (current line-up; earlier versions folded into their successors' lineage)",
"assets": 5,
"forecast_horizons": 5,
"cadence": "hourly",
"published_research_reports": 39
},
"source": "monthly_metrics_2026-09.json",
"control_reproduces_wave3_pin": null,
"pin_note": "Research-to-date counter: directional forecasts resolved inside the FH gate (1h/4h/1d) since Jul 11, counted at the issue cutoff (Thu Oct 1, 2026, 16:00 UTC) -- the D2 definition pinned in the weekly series; this issue prints 59,755. The pack carries no prior-cutoff pin for a monthly window (the prior-cutoff pin is null by construction), so the control this issue is the weekly reproduction: the W4 block of this pack (Sep 21-27) reproduces the published weekly #8 on 7 of 7 checks, and the prior window of this pack (Aug 1-31, prior cutoff Tue Sep 1, 2026, 16:00 UTC) reproduces the published Monthly #1 on 7 of 7 checks. Counter deltas against pre-wave-3 issues remain definitional.",
"control_last_week_reproduces_weekly8": {
"pins_source": "weekly_metrics_2026-09-21.json (the published weekly #8 pack)",
"note": "W4 of this run is the weekly #8 window with the weekly #8 frozen cutoff, so every pin must reproduce exactly. A mismatch means the window or the cutoff moved -- non-fatal here, but the monthly numbers must not be published until it is explained.",
"checks": {
"counters.mature_directional": {
"pin": 4958,
"got": 4958,
"match": true,
"tolerance": 0
},
"calibration.field.hit_rate": {
"pin": 0.4683,
"got": 0.4683,
"match": true,
"tolerance": 5e-05
},
"calibration.field.gap_pp": {
"pin": 16.1,
"got": 16.1,
"match": true,
"tolerance": 0.05
},
"consensus.loo_pooled.lift_pp": {
"pin": -14.0,
"got": -14.0,
"match": true,
"tolerance": 0.05
},
"model_watch.leaderboard.rows[0].model": {
"pin": "gemini-3.1-pro",
"got": "gemini-3.1-pro",
"match": true,
"tolerance": 0
},
"model_watch.leaderboard.rows[0].hit_rate": {
"pin": 0.5139,
"got": 0.5139,
"match": true,
"tolerance": 5e-05
},
"research_to_date.canonical_value_at_cutoff": {
"pin": 57929,
"got": 57929,
"match": true,
"tolerance": 0
}
},
"all_match": true,
"n_checks": 7,
"n_match": 7,
"last_week_key": "W4"
}
},
"site_entry": {
"slug": "consensus-watch-monthly-2026-09",
"series": "monthly",
"seriesLabel": "Monthly · Consensus",
"date": "October 2, 2026",
"title": "Monthly Consensus Watch #2",
"summary": "39,806 mature forecasts over one calendar month asked whether agreeing with the crowd makes an LLM market call safer. Herding ran at 99.2% of leave-one-out scoreable calls and the pooled monthly lift came in at -9.7pp: agree-hit 45.1% against 54.7% on disagreement, on a disagree bucket of n=137 and a 45.5% field base. The margin curve read 45.9% / 49.7% / 43.8% / 44.3% from 3 peers to full unanimity, while across the four weeks the weekly lifts ran -30.4pp to +12.1pp. Field accuracy moved 46.3% -> 45.5% (-0.8pp) against August, with 1 of 7 models improving.",
"stats": [
"17,625 peer-judged calls",
"99.2% ran with the herd",
"lift -9.7pp vs disagree"
],
"manifest_key_finding": "Herding ran at 99.2% of leave-one-out scoreable calls across September and the pooled lift came in at -9.7pp: agree-hit 45.1% against 54.7% on disagreement (disagree bucket n=137), against a 45.5% field base, while across the four weeks the weekly lifts ran -30.4pp to +12.1pp.",
"files": {
"pdf": "/research/reports/consensus-watch-monthly-2026-09.pdf",
"md": "/research/reports/consensus-watch-monthly-2026-09.md",
"json": "/research/reports/consensus-watch-monthly-2026-09.json"
}
},
"social": {
"telegram": {
"photo_caption_html": "🧭 Monthly Consensus Watch #2 — monthly series (window Sep 1–30)\n\nResearch question: if a model's forecast matches the consensus of the other models, is it more reliable?\n\n17,625 leave-one-out scored calls over one month, 7 model lines:\n• Herding ran at 99.2% — agree-hit 45.1% vs 54.7% when a model left the crowd (field base 45.5%, disagree n=137)\n• Margin curve: 3-peer 45.9%, 4-peer 49.7%, 5-peer 43.8%, unanimity 44.3%\n• Week by week the lift ran -30.4pp to +12.1pp across the four weeks\n• Market check: field accuracy 46.3% -> 45.5% (-0.8pp) vs August, 1 of 7 models improved; avg |1d| move 1.98% -> 1.91%\n\nNote: prior window = the full August (Aug 1-31), same rules as September\n\n📄 Full report (PDF) · all research\n\n#cryptotrading #LLM #AI #crypto #trading #consensus #research #Monthly",
"document_caption_html": "📄 Monthly Consensus Watch #2 — full report (PDF).\nWeb copy: consensus-watch-monthly-2026-09.pdf · machine-readable: JSON · all research\n\n#cryptotrading #LLM #AI #crypto #trading #consensus #research #Monthly",
"caption_len": 956,
"document_caption_len": 434,
"limit": 1024
},
"x": {
"main": "Monthly Consensus Watch #2 is out.\n\n17,625 peer-judged LLM calls over September: 99.2% ran with the herd, agreeing scored 45.1% vs 54.7% for the 137 that left it (-9.7pp).\n\nFull report: https://marketmania.ai/research/reports/consensus-watch-monthly-2026-09.pdf\n\n#AI #Crypto #Research #Consensus",
"main_tco_len": 243,
"reply": "Lift by week, then pooled:\nAug 31-Sep 6: -27.6pp\nSep 7-13: -30.4pp\nSep 14-20: +12.1pp\nSep 21-27: -14.0pp\nMonth: -9.7pp. Margin curve 45.9 / 49.7 / 43.8 / 44.3, field base 45.5.\nData: https://marketmania.ai/research/reports/consensus-watch-monthly-2026-09.json",
"reply_tco_len": 206,
"limit_tco": 280,
"replies_max": 1
}
},
"monthly_window": {
"start": "2026-09-01T00:00:00+00:00",
"end": "2026-10-01T00:00:00+00:00",
"text": "Sep 1-30, 2026 UTC",
"cutoff": "2026-10-01T16:00:00+00:00",
"cutoff_text": "Thu Oct 1, 2026, 16:00 UTC",
"days": 30,
"label": "Sep 1-30"
},
"prior_window": {
"start": "2026-08-01T00:00:00+00:00",
"end": "2026-09-01T00:00:00+00:00",
"text": "Aug 1-31, 2026",
"printed_as": "Aug 1-31",
"cutoff_text": "Tue Sep 1, 2026, 16:00 UTC",
"partial": false,
"note": "August is a full month: 2026-08-01 .. 2026-09-01, 31 days",
"applicable": true,
"not_applicable_reason": null,
"source": "month.prior_window -- window A of the same pull, matured at the prior cutoff with the same gate, maturity and hit rule",
"effective_start": "2026-08-01T00:01:00+00:00",
"counters": {
"forecasts_total": 42654,
"ok_in_gate": 40979,
"mature": 40979,
"mature_directional": 19383,
"mature_sideways": 21596
},
"definitions": {
"maturity": "horizon expired AND trade closed by the PRIOR cutoff (the same rule the current window uses at its own cutoff)",
"hit": "direction: exit tp1/tp2 -> hit, sl -> miss, expiry -> sign of gross pnl",
"fh_gate": [
"1h",
"4h",
"1d"
],
"min_n": 10,
"lineage": "rows spliced by the generalised 7-head lineage map before any aggregate",
"note": "identical definitions to the current window; only the window and the cutoff differ"
}
},
"weeks_index": {
"W1": {
"label": "Aug 31-Sep 6",
"issue": 5,
"published_slug_suffix": [
"2026-08-31T00:00:00Z",
"2026-09-07T00:00:00Z"
],
"bounds": [
"2026-08-31T00:00:00Z",
"2026-09-07T00:00:00Z"
],
"regime": "volatile"
},
"W2": {
"label": "Sep 7-13",
"issue": 6,
"published_slug_suffix": [
"2026-09-07T00:00:00Z",
"2026-09-14T00:00:00Z"
],
"bounds": [
"2026-09-07T00:00:00Z",
"2026-09-14T00:00:00Z"
],
"regime": "calm"
},
"W3": {
"label": "Sep 14-20",
"issue": 7,
"published_slug_suffix": [
"2026-09-14T00:00:00Z",
"2026-09-21T00:00:00Z"
],
"bounds": [
"2026-09-14T00:00:00Z",
"2026-09-21T00:00:00Z"
],
"regime": "volatile"
},
"W4": {
"label": "Sep 21-27",
"issue": 8,
"published_slug_suffix": [
"2026-09-21T00:00:00Z",
"2026-09-28T00:00:00Z"
],
"bounds": [
"2026-09-21T00:00:00Z",
"2026-09-28T00:00:00Z"
],
"regime": "calm"
}
},
"roster_rule": "7 models tracked (current line-up; earlier versions folded into their successors' lineage)",
"lineage_labels": {
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"spliced_in_window": [],
"cell_form": {},
"families": []
},
"controls": {
"last_week_reproduces_weekly8": {
"pins_source": "weekly_metrics_2026-09-21.json (the published weekly #8 pack)",
"note": "W4 of this run is the weekly #8 window with the weekly #8 frozen cutoff, so every pin must reproduce exactly. A mismatch means the window or the cutoff moved -- non-fatal here, but the monthly numbers must not be published until it is explained.",
"checks": {
"counters.mature_directional": {
"pin": 4958,
"got": 4958,
"match": true,
"tolerance": 0
},
"calibration.field.hit_rate": {
"pin": 0.4683,
"got": 0.4683,
"match": true,
"tolerance": 5e-05
},
"calibration.field.gap_pp": {
"pin": 16.1,
"got": 16.1,
"match": true,
"tolerance": 0.05
},
"consensus.loo_pooled.lift_pp": {
"pin": -14.0,
"got": -14.0,
"match": true,
"tolerance": 0.05
},
"model_watch.leaderboard.rows[0].model": {
"pin": "gemini-3.1-pro",
"got": "gemini-3.1-pro",
"match": true,
"tolerance": 0
},
"model_watch.leaderboard.rows[0].hit_rate": {
"pin": 0.5139,
"got": 0.5139,
"match": true,
"tolerance": 5e-05
},
"research_to_date.canonical_value_at_cutoff": {
"pin": 57929,
"got": 57929,
"match": true,
"tolerance": 0
}
},
"all_match": true,
"n_checks": 7,
"n_match": 7,
"last_week_key": "W4"
},
"month_vs_weeks": {
"month": 20682,
"sum_weeks": 19544,
"per_week": {
"W1": 4520,
"W2": 4468,
"W3": 5598,
"W4": 4958
},
"remainder_days": [
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"remainder_days_n": 3,
"remainder_mature_directional_month_pull": 1826,
"remainder_by_day": {
"2026-09-29": 668,
"2026-09-28": 772,
"2026-09-30": 386
},
"sum_weeks_plus_remainder": 21370,
"month_minus_sum_weeks_plus_remainder": -688,
"month_pull_by_week": {
"W1": 3832,
"W2": 4468,
"W3": 5598,
"W4": 4958
},
"note": "These WILL differ, and the difference is expected: each weekly pack is computed at its own frozen cutoff (week_end + 16h), while the MONTH pack is computed at 2026-10-01T16:00Z. A forecast still pending at its week's cutoff (trade not yet closed) is immature in the weekly pack and mature in the MONTH pack, so month >= sum_weeks + remainder. remainder_days_n counts the calendar days of the month outside the four Mon-Sun weeks (2026-09-28, 2026-09-29, 2026-09-30). month_pull_by_week is the same month pull split by slot week at the MONTH cutoff -- compare THAT with per_week to see the cutoff effect alone. W1 starts before the month (2026-08-31T00:00:00+00:00), so its pack also counts slots that are not in the MONTH pull and the inequality above need not hold for it."
},
"lineage_splice_counts": {
"MONTH": {
"claude-opus-4.8": 0,
"grok-4.5": 0,
"qwen-3.7-max": 0
},
"W1": {
"claude-opus-4.8": 0,
"grok-4.5": 0,
"qwen-3.7-max": 0
},
"W2": {
"claude-opus-4.8": 0,
"grok-4.5": 0,
"qwen-3.7-max": 0
},
"W3": {
"claude-opus-4.8": 0,
"grok-4.5": 0,
"qwen-3.7-max": 0
},
"W4": {
"claude-opus-4.8": 0,
"grok-4.5": 0,
"qwen-3.7-max": 0
}
},
"lineage_splice_detail": {
"MONTH": {
"total_window": 0,
"total_prev_window": 5390,
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"per_head": {
"claude-fable-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
},
"claude-opus-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
},
"deepseek-v4-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
},
"gemini-3.1-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
},
"gpt-5.6-sol": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
},
"grok-4.6": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
},
"qwen-3.8-max": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
}
}
},
"W1": {
"total_window": 0,
"total_prev_window": 110,
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"per_head": {
"claude-fable-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
},
"claude-opus-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
},
"deepseek-v4-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
},
"gemini-3.1-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
},
"gpt-5.6-sol": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
},
"grok-4.6": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
},
"qwen-3.8-max": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
}
}
},
"W2": {
"total_window": 0,
"total_prev_window": 0,
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"per_head": {
"claude-fable-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
},
"claude-opus-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
},
"deepseek-v4-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
},
"gemini-3.1-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
},
"gpt-5.6-sol": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
},
"grok-4.6": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
},
"qwen-3.8-max": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
}
}
},
"W3": {
"total_window": 0,
"total_prev_window": 0,
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"per_head": {
"claude-fable-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
},
"claude-opus-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
},
"deepseek-v4-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
},
"gemini-3.1-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
},
"gpt-5.6-sol": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
},
"grok-4.6": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
},
"qwen-3.8-max": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
}
}
},
"W4": {
"total_window": 0,
"total_prev_window": 0,
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"per_head": {
"claude-fable-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
},
"claude-opus-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
},
"deepseek-v4-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
},
"gemini-3.1-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
},
"gpt-5.6-sol": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
},
"grok-4.6": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
},
"qwen-3.8-max": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
}
}
}
},
"d2_at_cutoff": {
"window_key": "MONTH",
"cutoff": "2026-10-01T16:00:00+00:00",
"canonical_definition": "D2_directional_only",
"canonical_value_at_cutoff": 59755,
"since": "2026-07-11T00:00:00+00:00",
"d1_d3_at_cutoff": {
"D1_expired_ok_in_gate": 122919,
"D2_directional_only": 59755,
"D3_weekly_rule_mature": 122770
},
"marker": "RTD-D2-MONTHLY=59755",
"note": "this is the monthly RTD pin -- the number that goes into the Monthly plain boxes"
},
"prior_window_summary": {
"applicable": true,
"start": "2026-08-01T00:00:00+00:00",
"end": "2026-09-01T00:00:00+00:00",
"effective_start": "2026-08-01T00:01:00+00:00",
"cutoff": "2026-09-01T16:00:00+00:00",
"partial": false,
"note": "August is a full month: forecasts and candles from 2026-08-01",
"counters": {
"forecasts_total": 42654,
"ok_in_gate": 40979,
"mature": 40979,
"mature_directional": 19383,
"mature_sideways": 21596
},
"lineage_splice_counts": {
"claude-opus-4.8": 0,
"grok-4.5": 4520,
"qwen-3.7-max": 870
},
"field": {
"hit_rate": {
"month": 0.455,
"prior": 0.4633,
"delta_pp": -0.8
},
"gap_pp": {
"month": 17.0,
"prior": 16.3,
"delta_pp": 0.7
},
"brier": {
"month": 0.2796,
"prior": 0.2784
},
"mean_conf": {
"month": 62.6,
"prior": 62.6
},
"n": {
"month": 20682,
"prior": 19383
}
},
"consensus": {
"lift_pp": {
"month": -9.7,
"prior": 1.9,
"delta_pp": -11.6
},
"n_scoreable": {
"month": 17625,
"prior": 16089
},
"herd_share_pct": {
"month": 99.2,
"prior": 98.7
}
},
"leader": {
"month": "gemini-3.1-pro",
"month_hit_rate": 0.4851,
"prior": "claude-opus-5",
"prior_hit_rate": 0.4779
},
"per_model": {
"claude-fable-5": {
"hit_month": 0.4537,
"hit_prior": 0.4673,
"hit_delta_pp": -1.4,
"gap_month": 15.6,
"gap_prior": 13.9,
"gap_delta_pp": 1.7,
"n_month": 2945,
"n_prior": 2812
},
"claude-opus-5": {
"hit_month": 0.4538,
"hit_prior": 0.4779,
"hit_delta_pp": -2.4,
"gap_month": 15.9,
"gap_prior": 13.2,
"gap_delta_pp": 2.7,
"n_month": 2532,
"n_prior": 2281
},
"deepseek-v4-pro": {
"hit_month": 0.4526,
"hit_prior": 0.4554,
"hit_delta_pp": -0.3,
"gap_month": 15.3,
"gap_prior": 15.4,
"gap_delta_pp": -0.1,
"n_month": 3352,
"n_prior": 2578
},
"gemini-3.1-pro": {
"hit_month": 0.4851,
"hit_prior": 0.4761,
"hit_delta_pp": 0.9,
"gap_month": 18.0,
"gap_prior": 18.8,
"gap_delta_pp": -0.8,
"n_month": 3284,
"n_prior": 3033
},
"gpt-5.6-sol": {
"hit_month": 0.4446,
"hit_prior": 0.4586,
"hit_delta_pp": -1.4,
"gap_month": 23.5,
"gap_prior": 21.8,
"gap_delta_pp": 1.7,
"n_month": 3133,
"n_prior": 2970
},
"grok-4.6": {
"hit_month": 0.4439,
"hit_prior": 0.4527,
"hit_delta_pp": -0.9,
"gap_month": 15.8,
"gap_prior": 15.1,
"gap_delta_pp": 0.7,
"n_month": 2271,
"n_prior": 2708
},
"qwen-3.8-max": {
"hit_month": 0.4471,
"hit_prior": 0.4565,
"hit_delta_pp": -0.9,
"gap_month": 14.7,
"gap_prior": 14.7,
"gap_delta_pp": 0.0,
"n_month": 3165,
"n_prior": 3001
}
},
"per_symbol": {
"BNB": {
"hit_month": 0.4512,
"hit_prior": 0.4807,
"hit_delta_pp": -3.0,
"n_month": 3994,
"n_prior": 3857
},
"BTC": {
"hit_month": 0.4597,
"hit_prior": 0.4648,
"hit_delta_pp": -0.5,
"n_month": 3987,
"n_prior": 4068
},
"ETH": {
"hit_month": 0.4538,
"hit_prior": 0.4287,
"hit_delta_pp": 2.5,
"n_month": 3973,
"n_prior": 3695
},
"SOL": {
"hit_month": 0.4403,
"hit_prior": 0.4733,
"hit_delta_pp": -3.3,
"n_month": 4420,
"n_prior": 3928
},
"XRP": {
"hit_month": 0.4705,
"hit_prior": 0.4673,
"hit_delta_pp": 0.3,
"n_month": 4308,
"n_prior": 3835
}
},
"log_line": "prior_window: n_mature_directional=19383, effective_start=2026-08-01T00:01:00+00:00, field hit=0.4633 gap=16.3, loo lift=1.9",
"note_definitions": "the prior block is window A of the MONTH pull, matured at the prior cutoff with the same rules; delta = month - prior in percentage points"
},
"prior_reproduces_monthly1": {
"pins_source": "the published Monthly #1 JSONs calibration-monthly-2026-08.json, consensus-watch-monthly-2026-08.json, model-watch-monthly-2026-08.json (window 2026-08-01 .. 2026-09-01, cutoff 2026-09-01T16:00Z)",
"note": "month.prior_window of this run has the window, the cutoff and the definitions of the month Monthly #1 published; each check is printed and stored as a fact. A mismatch is not fatal and nothing is adjusted to match.",
"applicable": true,
"d2_cutoff_used": "2026-09-01T16:00:00+00:00",
"d2_slot_bound": "2026-09-01T00:00:00+00:00",
"checks": {
"prior_window.counters.mature_directional": {
"pin": 19383,
"got": 19383,
"match": true,
"tolerance": 0,
"published_path": "calibration-monthly-2026-08.json $.tables.counters.mature_directional"
},
"prior_window.calibration.field.hit_rate": {
"pin": 0.4633,
"got": 0.4633,
"match": true,
"tolerance": 5e-05,
"published_path": "calibration-monthly-2026-08.json $.tables.calibration_main[7].hit_rate"
},
"prior_window.calibration.field.gap_pp": {
"pin": 16.3,
"got": 16.3,
"match": true,
"tolerance": 0.05,
"published_path": "calibration-monthly-2026-08.json $.tables.calibration_main[7].gap_pp"
},
"prior_window.consensus.loo_pooled.lift_pp": {
"pin": 1.9,
"got": 1.9,
"match": true,
"tolerance": 0.05,
"published_path": "consensus-watch-monthly-2026-08.json $.tables.pooled.lift_pp"
},
"prior_window.model_watch.leaderboard[0].model": {
"pin": "claude-opus-5",
"got": "claude-opus-5",
"match": true,
"tolerance": 0,
"published_path": "model-watch-monthly-2026-08.json $.tables.leaderboard.rows[0].model"
},
"prior_window.model_watch.leaderboard[0].hit_rate": {
"pin": 0.4779,
"got": 0.4779,
"match": true,
"tolerance": 5e-05,
"published_path": "model-watch-monthly-2026-08.json $.tables.leaderboard.rows[0].hit_rate"
},
"research_to_date.at_prev_cutoff.D2_directional_only": {
"pin": 39073,
"got": 39073,
"match": true,
"tolerance": 0,
"published_path": "model-watch-monthly-2026-08.json $.research_to_date.platform.resolved_forecasts"
}
},
"all_match": true,
"n_checks": 7,
"n_match": 7
},
"note": "controls are DESCRIPTIVE and never fatal; they are printed in the log as well",
"builder_prior_vs_published_monthly1": {
"source": "month.prior_window of this pack vs the published Monthly #1 JSONs (calibration-monthly-2026-08.json, consensus-watch-monthly-2026-08.json, model-watch-monthly-2026-08.json)",
"applicable": true,
"checks": [
{
"check": "field hit rate",
"pack_prior_window": 0.4633,
"published_monthly1": 0.4633,
"tolerance": 5e-05,
"match": true,
"published_path": "calibration-monthly-2026-08.json tables.calibration_main[Field].hit_rate"
},
{
"check": "field gap pp",
"pack_prior_window": 16.3,
"published_monthly1": 16.3,
"tolerance": 0.05,
"match": true,
"published_path": "calibration-monthly-2026-08.json tables.calibration_main[Field].gap_pp"
},
{
"check": "field Brier",
"pack_prior_window": 0.2784,
"published_monthly1": 0.2784,
"tolerance": 5e-05,
"match": true,
"published_path": "calibration-monthly-2026-08.json tables.calibration_main[Field].brier"
},
{
"check": "LOO lift pp",
"pack_prior_window": 1.9,
"published_monthly1": 1.9,
"tolerance": 0.05,
"match": true,
"published_path": "consensus-watch-monthly-2026-08.json tables.pooled.lift_pp"
},
{
"check": "herd share pct",
"pack_prior_window": 98.7,
"published_monthly1": 98.7,
"tolerance": 0.05,
"match": true,
"published_path": "consensus-watch-monthly-2026-08.json tables.pooled agree.n / (agree.n + disagree.n)"
},
{
"check": "leader model",
"pack_prior_window": "claude-opus-5",
"published_monthly1": "claude-opus-5",
"tolerance": null,
"match": true,
"published_path": "model-watch-monthly-2026-08.json tables.leaderboard.rows[0].model"
},
{
"check": "leader hit rate",
"pack_prior_window": 0.4779,
"published_monthly1": 0.4779,
"tolerance": 5e-05,
"match": true,
"published_path": "model-watch-monthly-2026-08.json tables.leaderboard.rows[0].hit_rate"
},
{
"check": "mature directional",
"pack_prior_window": 19383,
"published_monthly1": 19383,
"tolerance": 0,
"match": true,
"published_path": "calibration-monthly-2026-08.json tables.counters.mature_directional"
}
],
"n_checks": 8,
"n_match": 8,
"all_match": true,
"note": "printed in the build log as facts; never a source of a printed number"
}
}
}